> I am writing an imap proxy. I want to limit the number of client
> connections. When the maximum limit is reached, I want to kill existing
> idle connections when a request for a new connection arrives.
This is a very bad design.
First, it provides a trivial denial-of-service attack. I can b
> According to last draft for IMAP server is allowed to return the folloing:
Yes, however today's IMAP4 *protocol* *standard* operates as I
described.
If and when a particular draft becomes an internet *standard* we
can then debate its characteristics. Internet standards based
applications are w
The IMAP4 specification is quite clear on this.
The text after "* OK" in the greeting is freeform (other than any
elements). The appearance of "IMAP4" or "IMAP4rev1"
there is meaningless when determining which version of IMAP4 is
supported.
You discover the level of IMAP4 the server supports by
> S: 354 go ahead
> C: blah blah...
> C: .
> S: 554 5.6.0 Message contains invalid header
LMTP wants RFC822 message body content in the DATA part. Your earlier
examples (output from ls getting piped into deliver) do not represent
well-formed RFC822 message content.
Do you get the same error if y
> You can check to see what Netscape (and any
> other IMAP Client) will use as a deliminator by logging into your IMAP
> server and sending the NAMESPACE command.
NAMESPACE is optional; not all servers support it. The portable method
for discovering the hierarchy delimiter is described in section
> Looks like this might have expired. Can't find it
Oops, you're right. There should be a new ACL draft issued just after the
Minneapolis IETF (if not sooner).
--lyndon
> I've brought this up several times before, but I'd like to offer it
> again. I think that the current ACL permission of CREATE for being able
> to delete folders is better than the old right of DELETE, but I think
> that it's still wrong.
The IMAP Extensions group in the IETF has a rework of A
> I heard/saw from somewhere that "the mailbox in cyrus imap server cannot
> be in a NFS filesystem, it only can be in local filesystem".
>
> Is this true? can somebody give me a reference pointer.
Yes. Cyrus requires rename() be an atomic operation. rename() is not
atomic on NFS filesystems.
S
